[Paper Review] Pricing Virtual Paths with Quality-of-Service Guarantees as Bundle Derivatives

Lars Rasmusson|ArXiv.org|Jun 12, 2001

Stochastic processes and financial applications9 references3 citations

TL;DR

This paper proposes a financial derivatives framework for pricing virtual paths with Quality-of-Service guarantees in computer networks by modeling network capacity as tradable assets. It uses a Girsanov transform to derive risk-neutral option pricing for multi-router paths, enabling continuous-time hedging independent of risk preferences, with prices computed via Monte Carlo simulation of correlated lognormal processes.

ABSTRACT

We describe a model of a communication network that allows us to price complex network services as financial derivative contracts based on the spot price of the capacity in individual routers. We prove a theorem of a Girsanov transform that is useful for pricing linear derivatives on underlying assets, which can be used to price many complex network services, and it is used to price an option that gives access to one of several virtual channels between two network nodes, during a specified future time interval. We give the continuous time hedging strategy, for which the option price is independent of the service providers attitude towards risk. The option price contains the density function of a sum of lognormal variables, which has to be evaluated numerically.

Motivation & Objective

• To enable end-users to trade complex network services as financial derivatives rather than relying on network-level admission control.
• To model network capacity in individual routers as underlying assets in a financial market, allowing dynamic, fine-grained resource negotiation.
• To develop a risk-neutral pricing model for options on virtual paths that span multiple routers, ensuring prices are independent of provider risk attitude.
• To design a continuous-time hedging strategy for network option writers using martingale techniques and stochastic calculus.
• To enable efficient market operation by reducing end-user negotiation overhead through derivative contracts on simple, fungible capacity shares.

Proposed method

• Models network capacity prices as correlated Itô processes with multiplicative drift, assuming mean-reverting behavior.
• Applies a novel Girsanov transform theorem to change the probability measure, enabling risk-neutral pricing of linear derivatives on path combinations.
• Derives a closed-form expression for the option price as a function of the sum of lognormal variables, requiring numerical evaluation via Monte Carlo simulation.
• Constructs a self-financing, continuous-time hedging strategy that replicates the option payoff and eliminates arbitrage opportunities.
• Uses the risk-neutral measure to simulate future capacity prices without tracking individual mean-reverting trajectories, improving computational efficiency.
• Introduces a derivative-based market structure where end-users trade options on virtual paths, with middlemen handling execution and pricing.

Experimental results

Research questions

• RQ1 How can complex network services like end-to-end virtual paths be priced as financial derivatives?
• RQ2 What mathematical framework enables risk-neutral pricing of options on multiple, correlated network resources?
• RQ3 Can a continuous-time hedging strategy be constructed that is independent of the network provider’s risk preferences?
• RQ4 How can the price of a path option be computed when it depends on the sum of lognormal-distributed router capacities?
• RQ5 What market structure enables efficient, low-overhead negotiation of complex network services by end-users?

Key findings

• The option price for a virtual path is given by a formula involving the expected value of a minimum cost path under a risk-neutral measure, with a correction term based on the probability that all paths exceed a strike price.
• The price depends on the density function of a sum of lognormal variables, which lacks a closed-form solution and must be evaluated numerically using Monte Carlo simulation.
• The Girsanov transform enables risk-neutral pricing by eliminating the drift from the underlying capacity processes, ensuring the option price is independent of risk attitude.
• The continuous-time hedging strategy is self-financing and replicates the option payoff exactly, eliminating arbitrage and ensuring market efficiency.
• The model supports a scalable market structure where simple capacity shares are traded on fast markets, and complex services are expressed as derivatives on these shares.
• The framework allows end-users to efficiently bid for or sell access to virtual paths without direct bilateral negotiation, reducing network overhead.